Xiaomi
Bonsai 27B
Bonsai 27B is an open-weights 27B-parameter model by Xiaomi. It holds an Intelligence score of 59 on the local.ai leaderboard (2026-08 snapshot).
open weights
59
local.ai Intelligence
Parameters
27B
Benchmarks (local.ai snapshot)
HuggingFace metadata · updated 2026-07-17
Metadata refreshed from the HuggingFace API (snapshot 2026-08-15T15:53:42.405Z). Context length and long-form descriptions are the next pipeline stage.
Does it fit your hardware?
Detect or select your GPU first. Green = weights + context headroom, yellow = tight, grey = does not fit.
Q4_K_M 15.7GQ6_K 24.3GQ8_0 29.7GFP16 54G| Quantization | ≈ File size | Note |
|---|---|---|
| IQ2_XXS | 7.3 GB | Extreme compression, largest quality loss |
| IQ3_XXS | 9.7 GB | Very small |
| Q3_K_M | 12.4 GB | Small |
| NVFP4 | 13.5 GB | NVIDIA 4-bit |
| Q4_K_S | 14.6 GB | Common |
| Q4_K_M | 15.7 GB | Most common sweet spot |
| Q5_K_M | 20.5 GB | Balanced |
| Q6_K | 24.3 GB | Near-lossless |
| Q8_0 | 29.7 GB | Near-lossless |
| FP16 | 54 GB | Original weights |
Actual files on HuggingFace
| File | Size |
|---|---|
| Bonsai-27B-F16.gguf | 50.11 GB |
| Bonsai-27B-dspark-bf16.gguf | 6.79 GB |
| Bonsai-27B-Q1_0.gguf | 3.54 GB |
| Bonsai-27B-dspark-Q4_1.gguf | 1.66 GB |
| Bonsai-27B-mmproj-BF16.gguf | 0.87 GB |
| Bonsai-27B-mmproj-Q8_0.gguf | 0.59 GB |
GGUF: prism-ml/Bonsai-27B-gguf · 1,801,896 downloads
GGUF: prism-ml/Ternary-Bonsai-27B-gguf · 775,667 downloads
| File | Size | |
|---|---|---|
| Ternary-Bonsai-27B-F16.gguf | 50.11 GB | download |
| Ternary-Bonsai-27B-PQ2_0.gguf | 6.67 GB | download |
| Ternary-Bonsai-27B-Q2_0.gguf | 6.67 GB | download |
| Ternary-Bonsai-27B-Q2_g64.gguf | 7.06 GB | download |
| Ternary-Bonsai-27B-dspark-Q4_1.gguf | 1.81 GB | download |
| Ternary-Bonsai-27B-dspark-bf16.gguf | 6.79 GB | download |
| Ternary-Bonsai-27B-mmproj-BF16.gguf | 0.87 GB | download |
| Ternary-Bonsai-27B-mmproj-Q8_0.gguf | 0.59 GB | download |
What you save self-hosting this model
Uses the model's own API price when it exists; otherwise the closest commercial equivalent.
Waiting for hardware detection…
Related models
RunLocal presents third-party leaderboard data (local.ai / Exo Labs, 2026-08) with its own fit and cost estimates. Verify pricing, licenses and file sizes at the source before deploying a model.